Skip to main content

High infection secondary attack rates of SARS-CoV-2 in Dutch households revealed by dense sampling

Our take —

The study sought to identify the secondary attack rate among households in Utrecht, the Netherlands, and whether the attack rate differed by age. They found that adults made up the plurality of cases (46% of 71 cases identified) and had the highest secondary attack rate, with 51% in adults compared to 41% in adolescents and 35% in children. The study was mostly made up of female healthcare workers as the index cases, and therefore may not be generalizable to other types of households that may not have medical knowledge or the possibility of healthcare-related exposure. Still, these data show a high overall risk of SARS-CoV-2 acquisition within the household of an infected case and a clear increase in secondary infection based on age.

Study design

Prospective Cohort

Study population and setting

This study sought to identify the secondary attack rate among households with a primary COVID-19 infection, as well as variability in the attack rate by age. An initial person who tested positive and was >18 years of age was reported by the Public Health Service in Utrecht, the Netherlands from March 24 to April 6 2020, and one household on May 24. The index case was contacted, and every household contact living with the index case enrolled in the study (except children <1 year of age). Within 24 hours after consent, household participants completed a questionnaire with demographic, medical and travel history, clinical symptoms, and treatment information. Participants reported if they had any symptoms two weeks before this first visit, and following this initial visit, they completed a daily symptom diary for two weeks. A second visit occurred two to three weeks after the first, and the final visit four to six weeks after the first. Participants were divided by age strata with adults 18+ years, adolescents 12 to 17 years, and children one to 11 years. Nasopharyngeal (NP), oropharyngeal (OP), oral fluid, and feces specimens were obtained. NP and OP swabs were combined to one result (negative if both were negative for SARS-CoV-2, or positive if either was positive) using polymerase chain reaction (PCR) testing. The secondary attack rate was calculated using generalized estimating equations (GEEs) that could account for household clustering. They also used a stochastic SEIR transmission model to estimate the secondary attack rate.

Summary of Main Findings

The study identified 55 index cases and 187 household contacts across 55 households. Household size varied from three to nine people. The majority of index cases were female (72.7%, N=40) and healthcare workers (75.9%, N=41). Among households, 17 (30.9%) had no additional transmission, while in 11 (20%), all individuals in the household were infected. Within the household, 50.7% (N=36/71) of adults, 45.6% (N=21/46) of adolescents, and 35.0% (N=21/60) of children were infected following the index case. The secondary attack rate was estimated at 43% (95% CI: 33% to 53%), and only age was significantly associated with increased likelihood of infection: the estimated secondary attack rate among children was 35% (95% CI: 24 – 46%), in adolescents it was 41% (95% CI: 27 to 56%), and in adults it was the highest at 51% (95% CI: 39 to 63%). The transmission rate was estimated at 1.2 transmissions per infection (e.g., each family member who is infected will likely infect 1.2 more people).

Study Strengths

The study had extensive follow-up in the cohort and was able to contact all household contacts, rather than only a subset who may be more likely to volunteer information or report infection. They also included children and adolescents, and were able to compare between these groups.


The study may be limited in its generalizability, as the majority of index cases were healthcare workers and testing at this time was primarily limited to symptomatic individuals. Therefore, this may not reflect transmission among asymptomatic index cases, or among households that may not have infection mitigation knowledge such as health workers would expectedly have. Additionally, index cases may not be the primary case, but rather just the first identified. Depending on how much further down the line they may have been infected may bias results towards increased secondary attack rates, as individuals who were infected last would be considered the index case for the entire family, even if transmission among those family members had occurred elsewhere, and not from the index case.

Value added

This study shows reduced secondary attack rate among children compared to adults, suggesting that children remain somewhat protected against SARS-CoV-2 from household transmission, though even here over one-third of children were infected.

This review was posted on: 14 May 2021